home *** CD-ROM | disk | FTP | other *** search
- Path: news.sfu.ca!samy
- From: samy@news.sfu.ca (Sam Yee)
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: TMapping again!
- Date: 19 Jan 1996 00:07:22 GMT
- Organization: Simon Fraser University
- Message-ID: <4dmnbq$pur@morgoth.sfu.ca>
- References: <mq2Xx*aB0@mkmk.in-chemnitz.de> <DLA1p3.A2n@si.hhs.nl>
- NNTP-Posting-Host: fraser.sfu.ca
- X-Newsreader: NN version 6.5.0 #5 (NOV)
-
- robv@si.hhs.nl (Veer) writes:
-
- >In article aB0@mkmk.in-chemnitz.de, floh@mkmk.in-chemnitz.de (Andre Weissflog) writes:
- >> In article <38232018@kone.fipnet.fi>, "Jyrki Saarinen" writes:
- >>
- >> Instead of 2 divs, one can also use 1 div and 2 muls:
- >>
- >> z' = 1 / one_div_z
- >> u = u_div_z * z'
- >> v = v_div_z * z'
- >>
-
- >This can even be done faster using a precalc table.
- >Remember: a*b = log(a+b).
-
- Oops! log(a*b)=log(a)+log(b)
-
- >Ofcourse, the size of the table can be minimize by examining the results of the log
- >table and the required precision. If the resulting values need not to be fixed-point
- >values between 0 and 512 the size of the table could be modified accordingly.
-
- >Though I have never tested this principle, I think a major speedup in calculations can
- >be achieved.
-
- >CYa,
-
- > Rob.
-
- >---
- >Think about this for a while:
-
- >Only fools can develop fool-proof applications.
-
-
-
-